基于知识的煮糖过程结晶状态检测分析研究  

Detection and analysis of sugar boiling crystallization state based on knowledge

作　　者：蒙艳玫[1] 何海平[1] 叶俊科[1] 王小纯[1] 唐治宏[1] 

机构地区：[1]广西大学机械工程学院,广西南宁530004

出　　处：《广西大学学报（自然科学版）》2013年第6期1357-1364,共8页Journal of Guangxi University（Natural Science Edition）

基　　金：国家自然科学基金资助项目(51065004);广西自然科学基金项目(2011GXNSFA018168);南宁市科技攻关项目(20131079)

摘　　要：针对煮糖过程机理异常复杂,包含的信息有定性的、定量的、半定量等多种模态的特点,结合专家知识和煮糖工人的经验,将粗糙集、支持向量机和专家系统有机结合在一起,构建基于知识的煮糖过程结晶状态分析专家系统。采用模糊C-均值聚类法对采集的样本进行离散化处理,基于粗糙集理论的启发式属性值约简,删除冗余信息,得到相应的结晶状态分析知识。为了提高煮糖过程结晶状态分析专家系统在线决策的能力,以获取的知识作为学习样本,构建基于支持向量机的煮糖过程结晶状态分析预测模型,仿真和实验结果证明模型是可行的,预测结果正确率为98%。Due to the extremely complicated mechanism in sugar boiling process with the multi-mode characteristics of qualitative, quantitative, and semi-quantitative information, an expert system of analyzing sugar boiling crystallization state based on knowledge is constructed by combining expert knowledge and worker experience with rough set, support vector machine and expert system. Fuzzy C-average clustering is applied to discriminate sample data. Based on attributive value reduction of rough set, redundancy information is reduced and analyzed knowledge corresponding to the crystalli- zation state is gained. In order to improve the expert system＇s online decision-making capacity, a prediction model was built to analyze the sugar boiling crystallization state by means of vector ma- chine and using the knowledge acquired as learning sample. The simulation and results of experi- ment show that the model is feasible with a correct prediction rate of 98%.

关 键 词：煮糖过程 粗糙集 专家系统 知识获取 支持向量机 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置]